The argument for automation has always contained an implicit promise. The hard, repetitive, dangerous work would be done by machines. Human beings, freed from the necessity of labor, would turn toward the things that actually mattered — creativity, relationship, contemplation, play.
This promise is old enough to predate the industrial revolution. It has never quite arrived. But the version being made now, about artificial intelligence and the end of cognitive labor, is more expansive than any previous iteration. And it deserves more skepticism than it is currently getting.
The historical record
The problem is not that automation fails to eliminate work. It does eliminate work — specific categories of it, in specific places, at specific times. The problem is what it replaces work with.
The historical record suggests the answer is rarely leisure in any meaningful sense. It is more often a different kind of work, performed under worse conditions, for less stable compensation, with less social recognition.
The cotton gin did not free enslaved people. The washing machine did not liberate women from domestic labor so much as raise the standards for cleanliness. The spreadsheet did not reduce the working hours of accountants.
The question is not whether machines can do our work. It is whether we know what we are without it.
There is something deeper here than economics. Work — even work that is tedious, even work that is poorly compensated — has functioned as a primary source of social identity, temporal structure, and a sense of competence for most people in most societies.
The psychological literature on unemployment is not primarily a literature about financial stress. It is a literature about the collapse of meaning. People do not just lose income when they lose work. They lose a reason to get up, a set of relationships, a role in the world that others can recognize and respond to.
Where automation actually lands
The communities most affected by industrial automation over the last forty years have not blossomed into centers of artistic production or civic engagement. They have, in many cases, experienced social fragmentation — addiction, declining health, diminished social trust — that follows from the removal of the structures that organize daily life.
The elimination of punishing work and the creation of meaningful alternatives are two separate projects. Treating the first as automatically producing the second is a category error with significant consequences.
Why this time is different
Previous automation replaced physical labor, then routine cognitive labor. The current wave reaches into domains — creative work, professional judgment, interpersonal skill — that have functioned as refuges for displaced workers and as sources of identity for people who did not work with their hands.
There are fewer obvious places to retreat to. The question of what human beings are for, in an economic sense, becomes genuinely open in a way it has not been before.
The end of work, if it comes, will not be the beginning of freedom by default. It will be the beginning of a question that we have not yet seriously tried to answer.